Metaplasia as documented in vertebrates (Steen, 1970; Yamada, 1967) is difficult to study in complex systems where cell transformations are not synchronous, many cell types are present, and cell movements complicate the task of tracing cell lineages. Thus it qould be preferable to study these transformations in an analogous system where complicating factors can be eliminated. Such a model system exists in the insect Manduca sexta (Lepidoptera). Here paired, tubular, simple epithelial glands function as silk glands for the caterpillar and as salivary glands for the adult moth. A single nondividing population of differentiated silk gland cells transform synchronously into four differentiated cell populations of the adult gland, without cell division, death, or large scale morphogenetic movements. Each adult population develops from a specific portion of the homogeneous larval population. Even the youngest (first instar) gland cells can transform to adult cells; thus larval cells possess the dual capacity to form either larval or adult differentiations. Using this system, the objective will be to study the timing and stability of adult determination. (1) Larval labial gland anlage will be transplanted from the embryo to new hormonal milieus to establish when determinations occur that specify the larval and adult fates. (2) Embryonic and larval transplantations will be used to study the specification of adult determined cells for a particular differentiation (of the four adult alternatives). (3) Finally, the stability of determination for the individual adult cell types will be tested through repeated passages through larval molts followed by metamorphosis; the age dependence of stability will be assessed by comparing the response of larval and adult glands to this procedure.